Andrew Blakers argues that complex energy models, like AEMO's Integrated System Plan, often produce narrow and misleadingly precise pathways for the grid. He points out that medium and long term forecasts fail because they cannot account for wars, politics, or technology cost shifts. A 2000-era model would have missed solar, batteries, and EVs entirely. Instead, he recommends high-speed simple models that map a broad range of good-enough energy futures rather than one false optimum. The piece uses a practical example: once error bars are added to a single optimal path, it expands into a wide landscape of equally valid options. That gives planners flexibility to adapt to real-world shocks. He also suggests normalizing solar, wind, and storage capacity against demand so results can be compared across regions. For anyone following Australian grid plans, this is a useful challenge to the authority given to big modelling exercises.
